Defending Jordan Peterson

Many people feel that the words they say are as much a part of their identity as their sexuality. We should go a long way to give people freedom of expression in both areas.

Jordan Peterson got in trouble with Twitter for a tweet saying:

Remember when pride was a sin? And Ellen Page just had her breasts removed by a criminal physician.

Jordan’s Twitter account is frozen unless he deletes the Tweet using a button that says “By clicking delete, you acknowledge that your Tweet violated the Twitter rules.” In a YouTube response, he said he “would rather die” than delete the tweet by his own action and doubled down with a critique of what is typically called “gender-affirming surgery,” performed on teenagers. His complaint with Elliot Page (formerly Ellen Page) is primarily that, as a role model (being a famous actor), Elliot might inspire more teenagers (mostly young women who would like to transition to being men) to deal with gender dysphoria by getting radical, dangerous, irreversible surgery. In his YouTube response to the Twitter ban, Jordan grudgingly allows that perhaps Elliot should have the right to undergo surgery to feel more like a man, but criticizes Elliot for encouraging others to do so.

My own experience with anyone who is transgender is limited to very positive experiences with Deirdre McCloskey. I don’t myself know what the right policies are in this area, though I tend to have Libertarian instincts, without being a full-fledged Libertarian. What I want to argue for, though, is that it is not only permissible, but essential, that our society have a vigorous debate about the appropriate age of consent for gender-affirming surgery. Parental input is another complicated issue, but I doubt anyone would say that a 6-year-old asking for gender-affirming surgery should be absolutely determinative, and let me write as if we agree that at age 30, anyone should be allowed to make that decision. Where should the line be drawn between 6 and 30? We default to age 18 on many things when faced with arbitrary determination of an age of consent. In this case, 18 is enough after puberty that it may cause technological problems to wait that long. But surely that makes figuring out good rules a societal decision that should be considered difficult rather than thinking it means we should short-circuit discussion by insisting peremptorily that technological concerns should trump concerns about the ability of teenagers to make wise irreversible decisions.

Let me make two analogies that suggest we should allow discussion about the age of consent for gender-confirming surgery—two analogies to suggest a range of views about this issue.

First, many feel plastic surgery for people who begin with a low normal appearance can be very helpful for their self-esteem and can really change their life for the better. Others fear that too many people are driven to get plastic surgery because of social pressure—or at least that this is true in some places, such as South Korea. Surely, people should be allowed to take either side in this debate!

Second, many (including me) feel that taking Psilocybin (the key ingredient in psychoactive mushrooms) can be very helpful to people in orienting their lives and may save the lives of a substantial fraction of those subject to suicidal ideation. Proponents of this view would like to see more states legalize the careful administration of Psilocybin, as Oregon has done. Others, for reasons I don’t fully understand, feel that clinical use of Psilocybin should not be allowed. Surely, however wrongheaded they are, we should not someone for arguing that Psilocybin should remain illegal.

If people should not be shouted down for expressing the view that the age of consent for gender-affirming surgery should be higher than it is, then the justification for requiring Jordan Peterson to delete his tweet in order to get his Twitter account unlocked becomes very thin.

Let me parse some of Jordan’s specific words.

“Sin”: Jordan loves the etymology of the word for sin in Greek. “Hamartia” means literally to miss the mark. So the claim here is that “Pride misses the mark.” Also, there is a well known Biblical verse: “Pride goes before destruction, a haughty spirit before a fall.” A reasonable interpretation of this part of Jordan’s tweet is therefore: “The current left-wing position and approach in the culture wars misses the mark and evidences a haughtiness that could lead to worrisome consequences for our society.”

“Ellen”: To some extent, Jordan’s use of “Ellen Page” to refer to Elliot Page is simply a reaffirmation that “The current left-wing position and approach in the culture wars—including especially its insistence on controlling other people’s speech—misses the mark and evidences a haughtiness that could lead to worrisome consequences for our society.”

Deadnaming, as in Jordan’s use of “Ellen” to refer to Elliot, is essentially a refusal to accept someone’s requests for how they should be referred to, plus perhaps a skepticism of that transgender transitions fully change someone’s gender. Skepticism that transgender transitions fully changes someone’s gender is a view that is likely shared by half of the American population. I don’t think that view should be beyond the pale. It is not denying anyone’s full humanity.

Stepping away from the transgender aspects of the situation, calling someone by a name they don’t like is extremely common in heated debates. For example, when Merrill Bateman was the President of Brigham Young University, punishing professors for espousing moderately liberal views, someone I knew referred to him in conversation as “Master Bateman.” (I was denied a job at BYU during this period—when I was a believing, temple-recommend-holding Associate Professor at the University of Michigan—for being too liberal.) Should (and does) Twitter reliably lock someone’s account for name calling? Or is it the transgression of the transgender orthodoxy that Twitter is responding to? Moreover, I think it would be a serious misreading of Jordan’s intent to think that he was trying to insult Elliot. His intent was to dispute the transgender orthodoxy. For Jordan, it wasn’t really about Elliot, except as an example of a broader issue.

“Her”: The things to say about Jordan’s use of the word “her” are almost identical to the issues with his use of “Ellen.”

“Criminal physician”: In his YouTube response, Jordan makes clear that he does, indeed, think that many of the gender-affirming surgeries that are performed should be outlawed. Just as important, and just as clear in his response, is that he thinks the social esteem in which surgeons who do such operations are held should be low. He is not advocating non-state violence or forcible action against such physicians, only laws restricting gender-affirming surgeries and social pressure against the number of gender-affirming surgeries being performed under the status quo.

Overall, there is no question that Jordan is emphatic in the expression of his views. But the views themselves and his expression of them seem well within the bounds of reasonable debate to me.

Let me say that, in his YouTube response, I think Jordan is too harsh about Elliot’s own decision, if it is possible to separate Elliot’s own decision from the charismatic example that it sets that might inspire and encourage many unhappy teenagers to make a female to male transition. In his 2017 Maps of Meaning lectures, Jordan emphasizes the ultimate responsibility of the individual to make decisions for their life, saying “What if no one knows any better than you?”:

In Jordan Peterson’s case, there are two reasons for me to defend him. First, freedom of speech is very dear to my heart. I left the Mormon Church because it was against freedom of speech within its orbit (church law, with excommunication and claimed suffering in the afterlife as the greatest penalty, not civil law). I am not about to cotton to abridgement of freedom of speech by a secular orthodoxy. (As an aside, there is an eerie similarity between the issue of deadnaming and the Mormon Church’s current insistent that it not be called “the Mormon Church,” but only by its full name of “The Church of Jesus Christ of Latter-day Saints.)

Second, Jordan Peterson is a brilliant exponent of important ideas. (I do not consider his views on transgender issues to be in that category. On the political Right, they are fairly conventional, unoriginal views.) Jordan being “controversial” in an area as charged as transgender issues could easily be enough to make many of my fellow academics stay away from reading or listening to Jordan Peterson entirely. That would be a mistake. In many areas—especially in the lectures and other long-form monologues in his YouTube videos—he is a deep and astute social commentator. (The short form of Twitter does not suit him well.) In his books 12 Rules for Life: An Antidote to Chaos and Beyond Order: 12 More Rules for Life, he gives advice that makes many people’s lives dramatically better. And in my view, he has gone further than anyone else in showing, in a practical sense, how to put the message of Christianity (and to some extent Judaism) on a nonsupernaturalist footing. That is an enormous achievement, in an area that I have taken as part of my own lifework. (To get a sense of my efforts in this area, see for example “How the Historical Jesus Set the Oppressed Free” and the sermons I have posted, including “Godless Religion,” “The Message of Jesus for Non-Supernaturalists” and “The Message of Mormonism for Atheists Who Want to Stay Atheists.”) Don’t disregard all of Jordan Peterson’s ideas because you are shocked or concerned by some! People who make other people think hard usually have at least one idea that is at serious variance with elite conventional wisdom. And in 2022, many feel protective of elite conventional wisdom at the “circle-the-wagons” level. Don’t let those forcefully protective of elite conventional wisdom scare you away.

Miles's Personality in 10 Facets of the Big Five

At understandmyself.com, Jordan Peterson offers a personality test for $10 based on an expansion of the Big Five personality dimensions into 10 aspects. The “Big Five” emerge from factor analysis of words in the language used to describe personality differences. To remember the Big Five, use either the acronym OCEAN or CANOE: Conscientiousness, Agreeableness, Neuroticism, Openness to Experience, Extraversion. The 10 aspects split each of this into two:

  • Conscientiousness is about Industriousness and Orderliness

  • Agreeableness is about Politeness and Compassion

  • Neuroticism is about Withdrawal and Volatility

  • Openness to Experience is about Intellect and “Openness” proper

  • Extraversion is about Assertiveness and Enthusiasm

Because factor analysis is maximizing how much can be accounted for by a given number of factors, the Big Five, or their expansion into 10 aspects, these factors account for a large share of the variance in personality. There isn’t that much left in personality differences between people that is not captured by the Big Five or these 10 aspects. Many interesting interpersonal differences with other labels can be reexpressed quite well as linear combinations of the Big Five or the 10 aspects from the Big Five. For example, being high on the two aspects of Openness to Experience and low on Orderliness is highly predictive of leaning left, politically; conversely, being high in Orderliness and low on Openness to Experience is predictive of leaning right, politically.

IQ is a different matter; it is not counted as personality, though it is often extremely valuable in predicting things that personality is also predictive of. For example, income is well predicted if you know IQ and Diligence. (The personality aspect of Intellect has to be carefully distinguished from IQ. Intellect is about being interested in abstractions. Some very smart people are interested mainly in concrete things; some people of average intelligence are fascinated by abstractions.) Intellectual abilities factor analyze with a very strong first factor. That means people good at one type of problem solving have a strong tendency to be good at each other type of problem solving. Howard Gardner made so much hay with his book Frames of Mind: The Theory of Multiple Intelligence because it takes a high IQ and great diligence to make a story that is so misleading in relation to the facts sound like it is on a solid footing. Of course, the first factor for intelligence does not exhaust the variance. But it is misleading not to lead with the fact that almost any two measures of intelligence have a high positive correlation. (That is, a high positive correlation in the population as a whole. Highly selective universities that choose all people with uniformly high intelligence in terms of the first factor of intelligence then have the remaining variance mainly from the smaller second and higher factors such as verbal vs. mathematical.)

As an economist, I am attracted to Principal Components Analysis, which fairly transparently reexpresses the variance-covariance matrix for a set of variables in terms of eigenvectors and their associated eigenvalues. In Principal Components Analysis, the different components (=factors) are uncorrelated by construction. But psychologists typical use factor analysis with rotation to get factors they consider easier to interpret. And these factors can have some correlation. In the case of the big 5, those correlations are on average .22. The two aspects within each of the Big 5 are reasonably distinct, with correlations within each pair between .35 and .59, as follows:

These are slides from one of the lectures up on YouTube for Jordan Peterson’s Personality class. They also illustrate that the Big Five can be divided into two groups: Plasticity and Stability. Note that “Emotional Stability” is simply Neuroticism times -1.

I should note that some of the other statements I make in this post are also drawn from having watched Jordan Peterson’s Personality class (as well as others of his classes) on YouTube.

Here are the results of my own personality test using the 10 aspects of the Big Five, expressed as percentiles:

  1. Agreeableness: 11

    • Compassion 42

    • Politeness 2

  2. Conscientiousness: 63

    • Industriousness: 97

    • Orderliness: 7

  3. Extraversion: 86

    • Enthusiasm: 79

    • Assertiveness: 85

  4. Neuroticism: 14

    • Withdrawal: 6

    • Volatility: 34

  5. Openness to Experience: 95

    • Intellect: 96

    • Openness: 84

The full report on me is here.

So I don’t sound too bad, let me quote this in the report about “Politeness”:

People who are exceptionally low in politeness challenge and confront authority – and they are not obedient. If they are respectful, it is grudgingly, and will only be manifested toward people who continually both deserve and demand it. They are comfortable confronting other people, and enjoy it. People extremely low in politeness are motivated to engage in conflict, and to seek out confrontation.

Note how different I am on the two aspects of agreeableness. I am also very different on the two aspects of conscientiousness: high on Industriousness, but low on Orderliness. In the description associated with being low in orderliness, I resonate to these two bits:

  • “… non-judgmental and devil-may-care in their attitudes toward themselves and others.”

  • “dislike schedules, list, or routines”

However, I like routines reasonably well if I made them up myself.

The aspects of the Big Five can help predict other outcomes of interest:

  • High IQ and Industriousness are good predictors of income. The other 9 aspects not so much. (Openness to Experience is positive correlated with creativity, but I’ll bet creativity is more closely related to the variance of income than to the mean of income.)

  • High Openness to Experience and Low Orderliness are good predictors of being politically Liberal.

  • High Agreeableness and High Orderliness are good predictors of insisting on political correctness. Conversely, Low Agreeableness and Low Orderliness (my bent) would predict a low level of political correctness.

Finally, being low on Withdrawal sounds like it is good for happiness. And indeed, I am generally quite happy. Here is some of what my report says about being very low on Withdrawal:

Individuals very low in withdrawal almost never suffer from or are impeded by anticipatory anxiety. They can handle new, uncertain, unexpected, threatening or complex situations very well. They are far less likely to avoid or withdraw in the face of the unknown and unexpected.

People with very low levels of withdrawal feel sad, lonesome, disappointed and grief-stricken very infrequently—and, if they do, do not feel those emotions deeply nor for long. Their lives tend to be markedly free of doubt, worry, embarrassment, self-consciousness and discouragement, even in the face of genuine threat and punishment. They are resistant to and rarely worried about social rejection, and almost never feel hurt or threatened.

Overall, I feel very lucky to have the disposition I have. People sometime act as if I am an alien, because I am extreme in some aspects of my personality. But I feel I have had a very enjoyable life.

To sum up, I am a hardworking, freewheeling, enthusiastic, assertive, happy, stable, convention-breaking artist-intellectual who cares about people.

With a Cobb-Douglas Production Function, the Differential Equation for the Solow Growth Model has a Closed-Form Solution—Tsering Sherpa and Miles Kimball

My “Ethics, Happiness and Choice” student (and research assistant) Tsering Sherpa did a project for her differential equation class on the Solow growth model. I was surprised to find out that the differential equation for the Solow Growth Model with a Cobb-Douglas production function has a closed-form solution. I proposed to Tsering that we coauthor a post on that. Here it is:


Economics put simply is the study of scarcity. The goal of economics is to study how resources are allocated, especially scarce resources. In order to do this, economists may use differential equations to model their analysis. A differential equation expresses the rate of change of the current state as a function of the current state, or in simpler terms, “an equation that involves an unknown function and its derivatives” (McOwen). More specifically, for the purposes of this post we look at ordinary differential equations. These are differential equations that only involve one independent variable. Ordinary differential equations relate to economics because they are the foundation for a multitude of mathematical concepts in economics. For example, the neoclassical growth theory introduced by Robert Solow and Trevor Swan in 1956 uses ordinary differential equations to describe how steady economic growth results from three main factors: labor, capital, and technology. This theory is also known as the Solow Growth Model.

The Solow Growth Model assumes capital and labor are being used efficiently, as well as that population growth, saving rate, and technology are constant. It also assumes a production function that is homogeneous of degree one: constant returns to scale. The goal is to solve for capital per worker, k, as a function of time. This shows exactly how capital per worker—and therefore the model economy—converges to its steady-state level, where the amount of capital lost by depreciation is offset by saving.

The differential equation driving the movements of capital per worker, k, is:

This equation gives the rate of change of capital per worker as a function of the fixed saving rate, s, the current level of capital per worker, k, output per worker, f(k), and the fixed depreciation rate, δ. The assumption of a Cobb-Douglas production function yields the more specific differential equation,

where 0< α < 1. The stipulation that 0< α represents capital being productive, while α < 1 represents the diminishing marginal returns to capital per worker. A is the rate of technological progress. The initial condition is

A change of variables reduces the differential equation (1) to a linear differential equation that can be solved by the usual integrating factor. Define

Then

The integrating factor is

Multiplying both sides by this integrating factor and rearranging:

Thus, using an indefinite integral:

where C is the constant of integration. The initial condition is:

Thus,

The steady-state value of capital per worker can then be seen as the limit of capital per worker as time goes to infinity:

The formula for capital per worker k is therefore a constant-elasticity-of-substitution (CES) aggregate between the initial level of capital per worker and the steady-state level of capital per worker k* given above, with elasticity of substitution equal to 1/α > 1 and a weight on the initial level of capital per worker that starts at 1 and exponentially decays at the rate (1-α)δ with the steady-state level of capital per worker k* having a complementary weight such that the two weights add to 1.

The formula for capital per worker, which drives all the other evolving variables in the model, implies that the convergence rate is equal to (1-α)δ. (That convergence rate generalize to cases with other production functions, as long as α is interpreted as capital’s share at the steady-state level of capital per worker.) This is a quite slow rate of convergence. For example, even if δ is relatively high, at a continuous-time rate of 10.5% per year, convergence would be a continuous-time rate of 7% per year if capital’s share is equal to 1/3. That means by the rule of 70 that the half-life of departures from the steady-state would be ten years, as the economy nears the steady state. (The rule of 70 is simply a consequence of the the natural logarithm of 2 equaling approximately .7.)

At the steady state, capital per worker is unchanging over time. That also means that unchanging at the steady state. Intuitively, investment is enough to compensate for depreciation. If there is population growth, or growth in the effective number of workers beyond population growth because of technological progress, the differential equation and its solution above continue to hold as long as k is interpreted as capital per effective worker and δ is interpreted as

δ = depreciation rate + population growth rate + rate of labor augmenting technological progress.

The power

is then the initial level of labor-augmenting technology, since labor-augmenting technology has to be taken to the power labor’s share (1-α) to be in multiplicative technology units, and conversely multiplicative technology level A has to be taken to the power of the reciprocal of labor’s share to be in labor-augmenting technology units. To see that, start with the two equivalent forms of the production function before the constant returns to scale has been used to express things in per effective worker terms. Here K is the overall level of capital, L is the overall number of workers and EL is the overall number of effective workers:

This equivalence implies

An interesting observation from the model is that two countries, regardless of the initial conditions for labor and capital, with the same savings, population growth, and depreciation rate will have the same steady state level of capital, thus both countries would conditionally converge (where the poor country grows faster along this convergence path).

Another important aspect of the Solow Growth Model is that interventions such as changing the saving rate will change the level of steady state per capita output, but will not permanently change the growth rate of per capita output.

In addition to its importance for economics, the Solow Growth Model provides a useful analogy for the effect of eating patterns on weight. As Miles discusses in “Kevin D. Hall and Juen Guo: Why it is So Hard to Lose Weight and So Hard to Keep it Off,” Permanent weight loss requires permanent changes in behavior, just as in the Solow Growth Model permanent changes in capital per effective worker require permanent changes in the saving rate.

Update, June 25, 2022, PM: Tweets from Dejanir Silva and Chris Edmond point out that Robert Solow himself showed this result and that this result was recapitulated by Chad Jones.

Update, June 29, 2022: Also see these tweets from Alex Zevelev for more growth models that can be solved in closed form.

References

  1. Acemoglu, Daron. MIT Economics. 1 Nov. 2011, https://economics.mit.edu/files/7181.

  2. Banton, Caroline. “The Neoclassical Growth Theory Explained.” Investopedia, Investope- dia, 19 May 2021, https://www.investopedia.com/terms/n/neoclassical-growth-theory.asp.

  3. McOwen, Robert C. Worldwide Differential Equations with Linear Algebra. Worldwide Center of Mathematics, LLC, 2012.

  4. “Solow Growth Model.” Corporate Finance Institute, 31 Jan. 2021, https://corporatefinanceinstitute.com/resources/knowledge/economics/solow-growth- model/.

  5. Whelan, Karl. Topic 1: The Solow Model of Economic Growth. Trinity College Dublin, 2005, https://www.tcd.ie/Economics/staff/whelanka/topic1.pdf.

Insufficient Sleep Contributes to Obesity, Diabetes, Heart Disease, Cancer and Neurodegenerative Disorders—Ken Wright

Here are some key quotations from the interview linked above, with bullets added to separate passages

  • … burning the candle all week and trying to “catch up on sleep” on the weekend not only doesn’t work well, but could actually worsen metabolic health in some ways.

  • … one easy way to recalibrate an off-kilter body clock, or circadian rhythm, is to go camping.

  • insufficient sleep contributes to all the major health problems, from obesity and Type 2 diabetes to heart disease, cancer and neurodegenerative disorders.

  • … toward a future in which, along with having their blood tested for cholesterol levels at the doctor’s office, patients might also get a test assessing whether they are natural night owls or morning larks and personalized prescriptions for how to better time their lives.

Price Stickiness Endangers Scientific Experiments Using Helium

There is a shortage of helium. But that shouldn’t be endangering high-value projects: a higher price can reduce the less important use of helium for party balloons or projects that can wait, while preserving the most essential uses of helium. But apparently there has been some reluctance to raise helium prices on the part of suppliers, resulting in rationing:

Vidoudez has spent countless hours calling or emailing just about every supplier that he could find.

“Most just either don’t answer or the ones that do answer say, ‘We don’t take new customers at the moment,’” Vidoudez said. “It’s been a real struggle.”

This is a good example of why those who raise prices when there are shortages are heroes, not the “price-gouging” villains they are sometimes made out to be. If you get a windfall because there is an unexpected shortage in what you sell, it would be great for you to give some of your windfall to charity, but keeping your price low is bad for society in such a situation.

Update, June 15, 2022: Of course, the “wealth effects” on scientific grants of fluctuations in helium prices can also be disruptive. A good answer to this, which physicists have pursued, is to have the equivalent of futures contracts for helium. Then the “wealth effects” of helium price fluctuations are zeroed out. Ideally, it is possible to take the money from the futures contract and not use the helium; that leaves the “substitution effect” (the incentive effect) of a price increase intact. And of course if the price goes down, the substitution effect is intact—one can buy more helium at the lower spot price if that is attractive.

Gradually Growing Sophistication in How Colleges are Evaluated

In “False Advertising for College is Pretty Much the Norm,” I argue:

Colleges and universities claim to do two things for their students: help them learn and help them get jobs. …

On job placement, the biggest deception by prestigious colleges and universities is to claim credit for the brainpower and work habits that students already had when they arrived. Another important deception is to obscure the difference between the job-placement accomplishments of students who graduate from technical fields such as economics, business, engineering or the life sciences, and the lesser success of students in the humanities, fields like communications or in most social sciences.

In Money’s “The Best Colleges in America, Ranked by Value,” the situation is getting a little better. A lot of Money’s focus is on providing possible students a more accurate idea of how much a college costs, which I didn’t focus on in my diatribe. But in some of the components of their evaluation described in their methodology explanation are in the direction I called for:

Major-adjusted early career earnings: (15%). Research shows that a student’s major has a significant influence on their earning potential. To account for this, we used College Scorecard program-level earnings of students who graduated in 2017 and 2018. These are the earnings of graduates from the same academic programs at the same colleges, one and two years after they earned a degree. We then calculated an average salary, weighted by the share of the bachelor’s degree completers in each major (grouped by CIP program levels). Then, we used a clustering algorithm to group colleges that have a similar mix of majors and compared the weighted average against colleges in the same group. Colleges whose weighted average was higher than that of their group scored well. This allows us to compare schools that produce graduates who go into fields with very different earnings potentials.

Value-added early career earnings (15%). We analyzed how students’ actual earnings 10 years after enrolling compared with the rate predicted of students at schools with similar student bodies.

Economic mobility: 10%. Think tank Third Way recently published new economic mobility data for each college, based on a college’s share of low- and moderate-income students and a college’s “Price-to-Earnings Premium” (PEP) for low-income students. The PEP measures the time it takes students to recoup their educational costs given the earnings boost they obtain by attending an institution. Low-income students were defined as those whose families make $30,000 or less. We used this as an indicator of which colleges are helping promote upward mobility. This data is new this year, and replaces an older mobility rate we had been using from Opportunity Insights.

Adjusting earnings for major and for the characteristics of students when they enter college is a step forward. I would like to have (a) seen these components emphasized more and (b) had the main table distinguish between earnings for different majors, so prospective students could see how much they sacrifice in earnings to pursue less-well-remunerated majors.

In “False Advertising for College is Pretty Much the Norm,” I also argue for direct measurement of learning:

To demonstrate that a school helps students learn, schools should have every student who takes a follow-up course take a test at the beginning of each semester on what they were supposed to have learned in the introductory course. The school can get students to take it seriously enough to get decent data — but not seriously enough to cram for it — by saying they have to pass it to graduate, but that they can always retake it in the unlikely event they don’t pass the first time. To me, it is a telling sign of how little most colleges and universities care as institutions about learning that so few have a systematic policy to measure long-run learning by low-stakes, follow-up tests at some distance in time after a course is over.

Some readers might argue that earnings are the ultimate goal, so measuring the intermediate outcome of learning is unnecessary. But I’ll bet that other unmeasured factors affect earnings to a greater degree than other unmeasured factors affect learning of the specific types of knowledge that colleges purport to teach. So measuring learning is a way to focus on an intermediate outcome likely to be to a greater extent causally due to college. And learning may be of some value to students in their lives beyond what it contributes to measured earning. For example, college graduates are healthier than those who don’t graduate from college, may make better consumer decisions and may get more enjoyment out of inexpensive entertainments such as reading books. Some colleges may do a better job at helping students learn things that will help their health, their decision-making and their enjoyment. As an even bigger deal, some colleges may do a better job at helping students choose life goals that those students will find meaningful in the future.

I describe some of my economics department’s effort to measuring learning in “Measuring Learning Outcomes from Getting an Economics Degree.” I discuss measurement of learning more in “How to Foster Transformative Innovation in Higher Education” and argue that measuring skill acquisition unlocks new, improved possibilities for higher education:

I want to make the radical claim that colleges and universities should, first and foremost, be in the business of educating students well. One implication of this radical claim is that colleges’ and universities’ performance should be measured by value added—by graduation rates and how much stronger graduating students are academically than they were at matriculation. By this standard, bringing in students who were impressive in high school raises the standards for what one should minimally expect a college’s or university’s students to look like at graduation, and colleges and universities become truly impressive if they help weak students become strong.

… The key to allowing alternative forms of higher education to flourish is to replace the current emphasis on accreditation, which tends to lock in the status quo, and instead have the government or a foundation with an interest in higher education develop high-quality assessment tools for what skills a student has at graduation. Distinct skills should be separately certified. The biggest emphasis should be on skills directly valuable in the labor market: writing, reading carefully, coding, the lesser computer and math skills needed to be a whiz with a spreadsheet, etc. But students should be able to get certified in every key skill that a college or university purports to teach. (Where what should be taught is disputed, as in the Humanities, there should be alternative certification routes, such as a certification in the use of Postmodernism and a separate certification for knowledge of what was conceived as the traditional canon 75 years ago. The nature of the assessment in each can be controlled by professors who believe in that particular school of thought.)

Having an assessment that allows a student to document a skill allows for innovation in how to get to that level of skill. …

To the extent colleges and universities claim to be teaching higher-order thinking, an assessment tool to test higher-order thinking is needed. One might object that testing higher-order thinking would be expensive, but it takes an awful lot of money to amount to all that much compared to four or more years of college tuition. And colleges and universities should be ashamed if they think we should take them seriously were they to claim that what they taught was so ineffable that it would be impossible for a student to demonstrate they had that skill in a structured test situation.

Conclusion: Outcome measurement matters. If you want to change the world for the better using your statistical or test-designing skills, figuring out how to do better outcome measurement for colleges and universities is a high leverage activity.

The Federalist Papers #55: How Big Should the House of Representatives Be?

The Federalist Papers #55 discusses the optimal size for the House of Representatives. The author (Alexander Hamilton or James Madison) argues that there is a large range of reasonable answers:

… no political problem is less susceptible of a precise solution than that which relates to the number most convenient for a representative legislature; nor is there any point on which the policy of the several States is more at variance, whether we compare their legislative assemblies directly with each other, or consider the proportions which they respectively bear to the number of their constituents.

Nevertheless, there are some concerns that could make the size of the House of Representatives too big or too small. The danger from being too big (besides the expense, which is not mentioned) is that it might act like a mob:

… the number ought at most to be kept within a certain limit, in order to avoid the confusion and intemperance of a multitude. In all very numerous assemblies, of whatever character composed, passion never fails to wrest the sceptre from reason.

Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob.

The behavior of large groups of people is something about which we have some social science evidence (synchronized movement can help submerge the individual within the group) and some experience that tends to bear out this concern (Occupy Wall Street and Black Lives Matter have gone to extremes because they have had no centralized leadership).

But it was the dangers of a too-small House of Representatives that were more on the minds of critics of the proposed Constitution. Quoting relevant passages, here are some of the concerns:

  1. so small a number of representatives will be an unsafe depositary of the public interests

  2. they will not possess a proper knowledge of the local circumstances of their numerous constituents

  3. they will be taken from that class of citizens which will sympathize least with the feelings of the mass of the people, and be most likely to aim at a permanent elevation of the few on the depression of the many

  4. it will be more and more disproportionate, by the increase of the people, and the obstacles which will prevent a correspondent increase of the representatives

The author of the Federalist Papers #55 allows that the first concern is important:

… a certain number at least seems to be necessary… to guard against too easy a combination for improper purposes

and adds a 5th reason for a body that is not too small:

5. a certain number at least seems to be necessary to secure the benefits of free consultation and discussion

It is taken as obvious that, at least numerically, 60 or 70 is enough to go a long way toward getting most of the benefit of having more heads to think something through. Answers to the 2d, 3d and 4th concerns is deferred to a later number.

On the worry that, if small in number, the House of Representatives could easily be bribed or could easily conspire, the Federalist Papers #55 argues that the number will soon be several hundred, and that the men who would be representatives when the body is initially less than one hundred had already passed the test of not trying to take tyrannical power during the Revolutionary War and after:

The Congress which conducted us through the Revolution was a less numerous body than their successors will be; they were not chosen by, nor responsible to, their fellow citizens at large; though appointed from year to year, and recallable at pleasure, they were generally continued for three years, and prior to the ratification of the federal articles, for a still longer term.

They held their consultations always under the veil of secrecy; they had the sole transaction of our affairs with foreign nations; through the whole course of the war they had the fate of their country more in their hands than it is to be hoped will ever be the case with our future representatives; and from the greatness of the prize at stake, and the eagerness of the party which lost it, it may well be supposed that the use of other means than force would not have been scrupled. Yet we know by happy experience that the public trust was not betrayed; nor has the purity of our public councils in this particular ever suffered, even from the whispers of calumny.

There is a strange discussion of whether the President and Senate would be rich enough to bribe the House of Representatives, motivated by the President and the Senate being a new element in the picture. But the key argument is the one above—that the likely members of the House of Representatives when it would be small were men who had been shown trustworthy by events.

How have things turned out? Now, the United States has 435 members of the House of Representatives. There are definitely concerns about those representatives being bought by moneyed interests, but I don’t see why those moneyed interests couldn’t work their co-opting magic on 4000 representatives. And I don’t notice a degeneration into acting like a mob from having so many as 435.

The question of how large a deliberative body should be is one that comes up all around the world, in many contexts. To mention just one, how many voting members should a monetary policy committee have?

Below is the full text of the Federalist Papers #55:


FEDERALIST NO. 55

The Total Number of the House of Representatives

From the New York Packet
Friday, February 15, 1788.

Author: Alexander Hamilton or James Madison

To the People of the State of New York:

THE number of which the House of Representatives is to consist, forms another and a very interesting point of view, under which this branch of the federal legislature may be contemplated.

Scarce any article, indeed, in the whole Constitution seems to be rendered more worthy of attention, by the weight of character and the apparent force of argument with which it has been assailed.

The charges exhibited against it are, first, that so small a number of representatives will be an unsafe depositary of the public interests; secondly, that they will not possess a proper knowledge of the local circumstances of their numerous constituents; thirdly, that they will be taken from that class of citizens which will sympathize least with the feelings of the mass of the people, and be most likely to aim at a permanent elevation of the few on the depression of the many; fourthly, that defective as the number will be in the first instance, it will be more and more disproportionate, by the increase of the people, and the obstacles which will prevent a correspondent increase of the representatives. In general it may be remarked on this subject, that no political problem is less susceptible of a precise solution than that which relates to the number most convenient for a representative legislature; nor is there any point on which the policy of the several States is more at variance, whether we compare their legislative assemblies directly with each other, or consider the proportions which they respectively bear to the number of their constituents. Passing over the difference between the smallest and largest States, as Delaware, whose most numerous branch consists of twenty-one representatives, and Massachusetts, where it amounts to between three and four hundred, a very considerable difference is observable among States nearly equal in population. The number of representatives in Pennsylvania is not more than one fifth of that in the State last mentioned. New York, whose population is to that of South Carolina as six to five, has little more than one third of the number of representatives. As great a disparity prevails between the States of Georgia and Delaware or Rhode Island. In Pennsylvania, the representatives do not bear a greater proportion to their constituents than of one for every four or five thousand. In Rhode Island, they bear a proportion of at least one for every thousand. And according to the constitution of Georgia, the proportion may be carried to one to every ten electors; and must unavoidably far exceed the proportion in any of the other States. Another general remark to be made is, that the ratio between the representatives and the people ought not to be the same where the latter are very numerous as where they are very few. Were the representatives in Virginia to be regulated by the standard in Rhode Island, they would, at this time, amount to between four and five hundred; and twenty or thirty years hence, to a thousand. On the other hand, the ratio of Pennsylvania, if applied to the State of Delaware, would reduce the representative assembly of the latter to seven or eight members. Nothing can be more fallacious than to found our political calculations on arithmetical principles. Sixty or seventy men may be more properly trusted with a given degree of power than six or seven. But it does not follow that six or seven hundred would be proportionably a better depositary. And if we carry on the supposition to six or seven thousand, the whole reasoning ought to be reversed. The truth is, that in all cases a certain number at least seems to be necessary to secure the benefits of free consultation and discussion, and to guard against too easy a combination for improper purposes; as, on the other hand, the number ought at most to be kept within a certain limit, in order to avoid the confusion and intemperance of a multitude. In all very numerous assemblies, of whatever character composed, passion never fails to wrest the sceptre from reason.

Had every Athenian citizen been a Socrates, every Athenian assembly would still have been a mob.

It is necessary also to recollect here the observations which were applied to the case of biennial elections. For the same reason that the limited powers of the Congress, and the control of the State legislatures, justify less frequent elections than the public safely might otherwise require, the members of the Congress need be less numerous than if they possessed the whole power of legislation, and were under no other than the ordinary restraints of other legislative bodies. With these general ideas in our mind, let us weigh the objections which have been stated against the number of members proposed for the House of Representatives. It is said, in the first place, that so small a number cannot be safely trusted with so much power. The number of which this branch of the legislature is to consist, at the outset of the government, will be sixty five. Within three years a census is to be taken, when the number may be augmented to one for every thirty thousand inhabitants; and within every successive period of ten years the census is to be renewed, and augmentations may continue to be made under the above limitation. It will not be thought an extravagant conjecture that the first census will, at the rate of one for every thirty thousand, raise the number of representatives to at least one hundred. Estimating the negroes in the proportion of three fifths, it can scarcely be doubted that the population of the United States will by that time, if it does not already, amount to three millions. At the expiration of twenty-five years, according to the computed rate of increase, the number of representatives will amount to two hundred, and of fifty years, to four hundred. This is a number which, I presume, will put an end to all fears arising from the smallness of the body. I take for granted here what I shall, in answering the fourth objection, hereafter show, that the number of representatives will be augmented from time to time in the manner provided by the Constitution. On a contrary supposition, I should admit the objection to have very great weight indeed. The true question to be decided then is, whether the smallness of the number, as a temporary regulation, be dangerous to the public liberty? Whether sixty-five members for a few years, and a hundred or two hundred for a few more, be a safe depositary for a limited and well-guarded power of legislating for the United States? I must own that I could not give a negative answer to this question, without first obliterating every impression which I have received with regard to the present genius of the people of America, the spirit which actuates the State legislatures, and the principles which are incorporated with the political character of every class of citizens I am unable to conceive that the people of America, in their present temper, or under any circumstances which can speedily happen, will choose, and every second year repeat the choice of, sixty-five or a hundred men who would be disposed to form and pursue a scheme of tyranny or treachery. I am unable to conceive that the State legislatures, which must feel so many motives to watch, and which possess so many means of counteracting, the federal legislature, would fail either to detect or to defeat a conspiracy of the latter against the liberties of their common constituents. I am equally unable to conceive that there are at this time, or can be in any short time, in the United States, any sixty-five or a hundred men capable of recommending themselves to the choice of the people at large, who would either desire or dare, within the short space of two years, to betray the solemn trust committed to them. What change of circumstances, time, and a fuller population of our country may produce, requires a prophetic spirit to declare, which makes no part of my pretensions. But judging from the circumstances now before us, and from the probable state of them within a moderate period of time, I must pronounce that the liberties of America cannot be unsafe in the number of hands proposed by the federal Constitution. From what quarter can the danger proceed? Are we afraid of foreign gold? If foreign gold could so easily corrupt our federal rulers and enable them to ensnare and betray their constituents, how has it happened that we are at this time a free and independent nation? The Congress which conducted us through the Revolution was a less numerous body than their successors will be; they were not chosen by, nor responsible to, their fellow citizens at large; though appointed from year to year, and recallable at pleasure, they were generally continued for three years, and prior to the ratification of the federal articles, for a still longer term.

They held their consultations always under the veil of secrecy; they had the sole transaction of our affairs with foreign nations; through the whole course of the war they had the fate of their country more in their hands than it is to be hoped will ever be the case with our future representatives; and from the greatness of the prize at stake, and the eagerness of the party which lost it, it may well be supposed that the use of other means than force would not have been scrupled. Yet we know by happy experience that the public trust was not betrayed; nor has the purity of our public councils in this particular ever suffered, even from the whispers of calumny. Is the danger apprehended from the other branches of the federal government?

But where are the means to be found by the President, or the Senate, or both? Their emoluments of office, it is to be presumed, will not, and without a previous corruption of the House of Representatives cannot, more than suffice for very different purposes; their private fortunes, as they must all be American citizens, cannot possibly be sources of danger. The only means, then, which they can possess, will be in the dispensation of appointments. Is it here that suspicion rests her charge? Sometimes we are told that this fund of corruption is to be exhausted by the President in subduing the virtue of the Senate. Now, the fidelity of the other House is to be the victim. The improbability of such a mercenary and perfidious combination of the several members of government, standing on as different foundations as republican principles will well admit, and at the same time accountable to the society over which they are placed, ought alone to quiet this apprehension. But, fortunately, the Constitution has provided a still further safeguard. The members of the Congress are rendered ineligible to any civil offices that may be created, or of which the emoluments may be increased, during the term of their election.

No offices therefore can be dealt out to the existing members but such as may become vacant by ordinary casualties: and to suppose that these would be sufficient to purchase the guardians of the people, selected by the people themselves, is to renounce every rule by which events ought to be calculated, and to substitute an indiscriminate and unbounded jealousy, with which all reasoning must be vain. The sincere friends of liberty, who give themselves up to the extravagancies of this passion, are not aware of the injury they do their own cause. As there is a degree of depravity in mankind which requires a certain degree of circumspection and distrust, so there are other qualities in human nature which justify a certain portion of esteem and confidence. Republican government presupposes the existence of these qualities in a higher degree than any other form. Were the pictures which have been drawn by the political jealousy of some among us faithful likenesses of the human character, the inference would be, that there is not sufficient virtue among men for self-government; and that nothing less than the chains of despotism can restrain them from destroying and devouring one another.

PUBLIUS.


Links to my other posts on The Federalist Papers so far:

Is Student Debt Forgiveness Fair?

Khaleda Rahman asked me for my views on cancelling existing student debt. Influenced by Richard Shinder’s Wall Street Journal op-ed shown below, I said we should make student debt dischargeable in bankruptcy.

Here is what I wrote to Khaleda in full (he used all but the last paragraph and the links):

Most Americans would view blanket student loan forgiveness as unfair to those who sacrificed to pay off their loans. And the vast majority of college students come from the upper half of the income distribution. We already have a system for loan forgiveness for those who are in dire financial trouble: it is called bankruptcy court. We should make student loans eligible to be discharged or modified in bankruptcy on the same basis as other loans. As it is now, they can't be discharged in bankruptcy.

Part of the problem students have in paying off loans is not the loans themselves, or even the high cost of college more generally, but that often students aren't getting a good education, or aren't given a true picture of their financial prospects after different majors. Colleges and universities need to have their feet held to the fire to do collect data and do honest reporting about the quality of their education and the financial prospects of students who follow different tracks. I write about this in my Bloomberg piece "False Advertising for College is Pretty Much the Norm.”

On the high cost of college, disruptive innovation that can dramatically reduce the cost of a college education already has a foot in the door. We just need to welcome it in. I write about that in my Quartz column “The Coming Transformation of Education: Degrees Won’t Matter Anymore, Skills Will.”

A Flexiblog

As I said in the 10th anniversary post for my blog, “Everything is Changing.” I have been realizing that one of the things I love most is to be able to do tasks and projects in whatever order feels right to me at the time. Having to do things in order of urgency is sometimes necessary, but feels like a cost to me. So I’m going to try an experiment of blogging with less of a weekly schedule to my posts. I plan to write as I feel inspired to write, aiming to continue to write at a good pace overall.

In some ways, this is a return to how I did my blog at the very beginning. Then I had no particular pattern to my posts. They came out whenever I had an idea. (Of course, then I was so excited by beginning to blog that I couldn’t sleep!)

There is a Golden Mean between being too flexible and being too rigid. (See “The Golden Mean as Concavity of Objective Functions.) I think I have been too focused on avoiding the harm from being too flexible (thinking of it as being undisciplined) and not focused enough on the harm from being too rigid. But I’ll try not to overcorrect.

Everything is Changing

Today is the 10th anniversary of this blog, "Confessions of a Supply-Side Liberal." My first post, "What is a Supply-Side Liberal?" appeared on May 28, 2012. I have written an anniversary post every year since then:

  1. A Year in the Life of a Supply-Side Liberal

  2. Three Revolutions

  3. Beacons

  4. Why I Blog

  5. My Objective Function

  6. A Barycentric Autobiography

  7. Crafting Simple, Accurate Messages about Complex Problems

  8. On Human Potential

  9. Pandemic Passage: My Past 12 Months in Blogging

I don’t say lightly that I feel more than ever in my adult life that the world is changing. The pandemic, besides its direct stresses, has made us all look at work differently. Large-scale war has returned to Europe. Political polarization and associated bad behavior is worse than it has been since the 1960s.

More parochially, the audience for blogs seems to be changing. I feel I am serving a different readership than I was a few years ago, as many are drawn off toward following the news of the other changes I mentioned above. I can’t predict where things will go in the future.

In my personal life, everything seems different in 2022. New Year’s day found me and my family temporary refugees from a wildfire that destroyed about 10% of the houses in my town of Superior, Colorado and the neighboring city of Louisville, Colorado. (See “New Year's Gratitude on the Occasion of the Marshall Fire.”) Soon after that I began a semester in which I put a concerted effort into creating a new course and improving an existing course. (See “Ethics, Happiness and Choice—Miles's Economics 4060,” “Intermediate Macroeconomics—Miles's Economics 3080.”) It felt more like I was changing than simply what I was teaching. On May 15, our youngest, Jordan, married Caroline. That, too, has far-reaching ramifications us as a family as well as for them as a couple. (Our daughter Diana married Erik in 2017. Jordan’s marriage to Caroline means both of our children are now married.)

To handle everything going on (my research continued at full speed), for the first time in a long time I missed doing some blog posts in my usual 3-times-a-week schedule (not counting link posts). But as I look back over the year’s blog posts using the “Archive” button up above I am amazed at how well I did manage to keep up the pace on my blog.

For the past decade, blogging has been a major part of my life. It gives meaning to every week as I put down in words what I have been learning and thinking and try to influence in some small way the path our civilization is taking.

A decade from now, I plan to retire. I think I can do a lot in that time, personally, academically and on this blog. And I have big plans for my retirement after that, beginning with writing an autobiography. 

I look forward to seeing where the world will go in the next ten years. I am an optimist. Event sometimes thrust us into the underworld, but we learn things there and with a little luck and a lot of fortitude, we can come back stronger and more true to our deepest values. May we all strive toward a better world ten years from now than the one we see around us now. I’ll meet you there.

Cognitive Behavioral Therapy for Insomnia Can Prevent Major Depression

Why do we care about causality? The primary reason is that we are looking for interventions that can make things better. Some causal results give only one part of the puzzle of designing an intervention to make things better, while other causal results directly recommend an intervention once combined with a reasonable judgment about what constitutes “better.”

The result in “Prevention of Incident and Recurrent Major Depression in Older Adults With Insomnia” by Michael R. Irwin, Carmen Carrillo, Nina Sadeghi, Martin F. Bjurstrom, Elizabeth C. Breen, and Richard Olmstead that “Cognitive Behavioral Therapy for Insomnia” (CBT-I) applied to adults 60 years or older who have insomnia, but don’t initially qualify as having major depression both reduces insomnia and reduces the incidence of major depression. This is a very useful thing to know.

It is tempting to interpret this result as strong evidence that insomnia causes an increased incidence of major depression. It is quite plausible that is true, since many experiments show that extreme sleep deprivation (or extreme deprivation of just REM—dreaming—sleep) quickly causes people to go crazy. But underlying subclinical depression could easily disrupt sleep, so there is a real possibility of reverse causality. And as a statistical instrument, CBT-I does not satisfy the exclusion restriction: Cognitive Behavioral Therapy (CBT) in general is known to have a powerful effect in reducing depression even in people who don’t have insomnia. So the CBT elements in CBT-I are likely to have a direct effect in reducing depression that doesn’t all work through reducing insomnia.

What about fact that those whose insomnia lessened were especially likely to have a reduction in incidence of major depression? I am thinking of this result:

Those in the CBT-I group with sustained remission of insomnia disorder had an 82.6% decreased likelihood of depression (hazard ratio, 0.17; 95%, CI 0.04-0.73; P = .02) compared with those in the SET group without sustained remission of insomnia disorder.

That can be explained simply by lessened insomnia being a sign that the CBT principles “took” for that individual. That is, CBT is going to work better for some people than others, if only because after being taught CBT principles, some people will really use those principles in their lives and others won’t. If CBT “takes” for an individual, it can have benefits in multiple areas.

Should we care whether reduction in insomnia is the pathway by which CBT-I reduces the incidence of depression? Not if CBT-I is, and would stay, the only effective treatment for insomnia. But when there is another effective treatment for insomnia, it matters. Whether that other treatment for insomnia will reduce the incidence of depression depends on the extent to which the pathway by which CBT-I reduce depression is through lessening insomnia. The best way to find that out is to do a similar study with other effective treatments for insomnia. If enough different effective treatments for insomnia also reduce the incidence of depression, that raises confidence that the next effective treatment for insomnia will also reduce the incidence of depression, which is the most important practical thing we would mean by claiming that insomnia causes a higher incidence of depression.

Statistically, many different effective treatments for insomnia reducing the incidence of depression would make it more likely that at least one of those treatments approximately satisfied the exclusion restriction by not having a big direct effect on incidence of depression. Note how that would depend on how different the “different effective treatments for insomnia” are. For example, in the extreme, if they were all minor variations on CBT-I, then they don’t lend much additional evidentiary weight to the claim that insomnia causes a higher incidence of depression.